Thinkcomp, I agree with almost everything you say about the economics of the legal profession, but I think you miss out on something crucial.
Before I went to law school, I unconsciously thought of law as a branch of engineering: Learn the rules, twist the right dials, and things will happen as you expect. That was naive. Law does have rules, and dials to twist, but for socio-political reasons I won't go into here, the rules and dials aren't tightly locked to outcomes. As a result, law has long struck me as being more akin to weather-guessing than to engineering.
The article "The Myth of the Rule of Law"[1] has been linked a few times on HN, and I think it sums up this idea pretty well. It basically says that in many (most?) cases the law is subject to interpretation that depends on the politics of the interpreter (be they a lawyer, judge or member of a jury). The "letter of the law" isn't nearly as simple as people believe, ambiguity is the name of the game.
As an engineer, that's why I went to law school. I thought law could be represented deterministically...but that's a naive assumption. A stochastic model fits much better, which is why I'm now extremely interested in artificial intelligence.
Before I went to law school, I unconsciously thought of law as a branch of engineering: Learn the rules, twist the right dials, and things will happen as you expect. That was naive. Law does have rules, and dials to twist, but for socio-political reasons I won't go into here, the rules and dials aren't tightly locked to outcomes. As a result, law has long struck me as being more akin to weather-guessing than to engineering.